APPLIED AND ENVIRONMENTAL MICROBIOLOGY, OCt. 1994, p. 3854-3861 Vol. 60, No. 100099-2240/94/$04.00+0Copyright © 1994, American Society for Microbiology

Sensitization of Heat-Treated Listeria monocytogenes toAdded Lysozyme in Milk

DAVID J. KIHM,1t GREGORY J. LEYER,1' GIL-HWAN AN,'§ AND ERIC A. JOHNSON' 2*Departments of Food Microbiology and Toxicology' and Bacteriology, 2

University of Wisconsin, Madison, Wisconsin 53706

Received 31 March 1994/Accepted 21 July 1994

Listeria monocytogenes was highly resistant to hen egg white lysozyme in whole milk but was sensitive inmedia and in phosphate buffer. Methods to sensitize the pathogen to lysozyme in milk were investigated.Treatment of whole milk by cation exchange to remove minerals, particularly Ca2' and Mg2+, slightlypromoted inactivation of L. monocytogenes by lysozyme at 4°C over a period of 6 days. Heat treatment (62.5°Cfor 15 s) strongly sensitized L. monocytogenes to lysozyme in demineralized milk and in MES [2-(N-morpholino)ethanesulfonic acid] buffer. Addition of Ca2+ or Mg2+ to the demineralized milk restoredresistance to lysozyme. Cells were more rapidly heat inactivated at 55°C in demineralized milk containinglysozyme, and addition of Ca2+ to the demineralized milk restored the resistance to heat. The results indicatethat minerals or mineral-associated components protect L. monocytogenes from inactivation by lysozyme andheat in milk, probably by increasing cell surface stability. The heat treatment of foods containing addedlysozyme can probably play a significant role in producing microbiologically safe foods.

During the past decade, certain dairy products have beenassociated with major outbreaks of listeriosis (5, 11, 30).Listeria monocytogenes continues to be a serious concern inregard to food safety because of the high fatality rate of about30% in infected humans and the high susceptibility of certaingroups of humans (30). Humans susceptible to listeriosisinclude immunocompromised individuals with underlying dis-eases and pregnant women and their developing fetuses. Theseindividuals should avoid eating foods that may be contami-nated with the pathogen.

Avoiding contamination of many raw and minimally pro-cessed foods by L. monocytogenes is a difficult task for the foodindustry. L. monocytogenes is widespread in nature and hasbeen isolated from soil, decaying vegetation (including silage),and human and animal feces. It has also been found in dairyfoods, including raw milk and some ripened cheeses (28). L.monocytogenes has several phenotypic characteristics that maypromote growth and survival in dairy products, includinggrowth over the temperature range of 1 to 45°C, tolerance to>10% sodium chloride, and growth at pH 5.5 to 9.6 (31).Because of its pathogenicity and resistance properties, L.monocytogenes has been considered a relatively high-riskpathogen for the dairy industry (19).

Because of the ubiquity of L. monocytogenes and its rela-tively frequent contamination of raw and minimally processedfoods, considerable attention has been given to developingantimicrobial agents that would kill the pathogen in theseproducts. Our laboratory showed that egg white lysozymeinactivated L. monocytogenes in microbiological media and invarious foods (16, 17). Inactivation of L. monocytogenes bylysozyme was highly dependent on the food system. L. mono-cytogenes was killed by lysozyme in vegetables, but it wasrelatively resistant in foods of animal origin, including milk and

* Corresponding author. Mailing address: Department of FoodMicrobiology and Toxicology, University of Wisconsin, 1925 WillowDr., Madison, WI 53706. Phone: (608) 263-7944. Fax: (608) 263-1114.

t Present address: Abbott Laboratories, North Chicago, IL 60064.t Present address: Ross Laboratories, Columbus, OH 43219.§ Present address: Hai-tai Confectionary Co., Seoul, South Korea.

Camembert cheese (17). Further studies by us and by others(8) have shown that L. monocytogenes is strongly resistant tolysozyme in milk. No lysozyme activity against L. monocyto-genes strains in whole milk was found, whereas killing in acidmilk (pH 5.3) was detected (8). We undertook this study todetermine the basis for this resistance and to develop methodsto sensitize the pathogen to lysozyme in milk.

MATERIALS AND METHODS

Materials. The sources of egg white lysozyme hydrochlorideand Micrococcus luteus were the same as previously described(17). Chemical reagents were of the purest grade available andwere obtained from Sigma Chemical Co., St. Louis, Mo.Microbiological media were obtained from Difco Laborato-ries, Inc., Detroit, Mich. Whole pasteurized milk was pur-chased from the Department of Food Science dairy store,University of Wisconsin, Madison. Ultrapure water was pre-pared by passing distilled water through a Nanopure II mixed-bed ion-exchange system (Barnstead Co., Newton, Mass.). Allreagents and buffers were prepared with ultrapure water.

Bacterial strains and culture conditions. L. monocytogenesScott A was originally provided by E. H. Marth, University ofWisconsin, Madison. L. monocytogenes was grown in staticculture in tryptic soy broth [TSB] at 37°C and maintained onTSB agar plates at 4°C with monthly transfers. Growth inculture media was measured by optical density at 500 nm in aSpectronic 20D spectrophotometer (Milton Roy Co., Roches-ter, N.Y.).

L. monocytogenes was enumerated after exposure to ly-sozyme or heat by plating on tryptose-phosphate agar with 1%(wt/vol) sodium pyruvate (TPAP), because this medium effec-tively recovers healthy as well as heat-injured cells (32). TPAPcontained, per liter, 20 g of tryptose, 2.0 g of glucose, 2.5 g ofyeast extract, 2.5 g of NaCl, 2.5 g of Na2HPO4 7H20, 10 g ofsodium pyruvate, and 15 g of agar. Plates were incubated at37°C for 2 days.Heat treatment of L. monocytogenes. To determine the

susceptibility of heat-treated cells to lysozyme in milk, L.monocytogenes was grown statically overnight in TSB, 1 ml was

3854

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 30

Dec

embe

r 20

21 b

y 14

4.13

2.16

8.21

.

INACTIVATION OF L. MONOCYTOGENES IN MILK BY LYSOZYME 3855

.

0

10

031 02

0 50 100 150 200

Time (hours)FIG. 1. Survival of L. monocytogenes in the presence of lysozyme

(100 mg/liter) in whole milk and demineralized whole milk at 4°C.Symbols: 0, unheated L. monocytogenes in whole milk with or withoutlysozyme; 0, unheated or heat-treated (60°C for 15 s) cells indemineralized whole milk without lysozyme; O, heat-treated cells inwhole milk with or without lysozyme; M, unheated cells in demineral-ized whole milk containing lysozyme; A, heat-treated L. monocyto-genes in demineralized milk containing lysozyme; A, unheated cells in67 mM PB, pH 6.6, containing lysozyme.

harvested by centrifugation, and the cells washed in 1 ml of 67mM sodium phosphate buffer (PB), pH 6.4, and resuspendedin 100 ,ul of PB. The 100-,ul cell suspension was added to 4.9 mlof PB at the desired temperature, and the suspension wasmixed by vortexing and incubated for 13 s. The heated cellsuspension was then transferred to 20 ml of ice-cold PB andmixed rapidly for immediate temperature cooling, and the cellswere harvested by centrifugation. With the aid of an assistant,it was possible to transfer from heating menstrum to coldbuffer in 2 s, resulting in a 15-s total heat treatment. These cellswere used for inoculation to milk or buffer, which was incu-bated at 4°C. Survival was determined by plating on TPAP.

Resistance of L. monocytogenes to lysozyme in milk A100-fold stock solution of lysozyme hydrochloride was dis-solved in PB, filter sterilized, and added to obtain a finalconcentration of 100 mg/liter of milk. L. monocytogenes wasinoculated into the milk, samples were held at 4°C, andsurvivors were enumerated by plating on TPAP. Metals(CaCl2, MgCl2, and ZnCl2) were dissolved in distilled water,filter sterilized, and added back to demineralized milk toobtain concentrations similar to those found in whole milk.Metal analyses of whole milk, demineralized milk, and refor-tified milk were performed at the Soil and Plant AnalysisLaboratory, University of Wisconsin, Madison, by using induc-tively coupled plasma spectroscopy as previously described forspores of Clostridium botulinum (20).

TABLE 1. Metal analysis of demineralized milk samples

Concn (mg/liter) in:Metal Dialyzed milk +

Whole milk Dialyzed milk Chelex'

P 979 563 474K 1,839 <7 <7Ca 1,170 325 100Mg 118 21 <1S 285 195 157Zn 3.5 1.5 <0.1Fe <0.13 <0.13 <0.13Cu <0.29 <0.28 <0.28Na 448 773 899

a Chelex was used at a concentration of 50 g/100 ml of milk.

Resistance of L. monocytogenes to lysozyme in bufer. Thecapacity of metals to restore resistance to lysozyme wasevaluated by adding washed cells previously heat treated in PBto 5 ml of 50 mM MES [2-(N-morpholino)ethanesulfonicacid]) buffer, pH 6.4, containing 100 mg of lysozyme per liter.CaCl2 and MgCl2 were added to the buffer at the same levelsas in whole milk (1,100 mg/liter for Ca2' and 110 mg/liter forMg2+). The suspensions were held at 4°C, and viability overtime was determined by plating on TPAP.

Demineralization of whole millk To remove metals fromwhole milk, the pH of the milk was adjusted to 6.2 withconcentrated citric acid, and the milk was dialyzed overnight at4°C in two changes of 10 mM EGTA [ethylene glycol-bis(p-aminoethyl ether)-N,N,N',N'-tetraacetic acid], pH 6.4, usingdialysis tubing with a molecular weight cutoff of 6,000 to 8,000(Spectrum Medical Industries, Inc., Los Angeles, Calif.). Themilk was dialyzed the next day against two changes of 20 mMsodium phosphate buffer, pH 6.4. Certain batches of milk werefurther treated with Chelex 100 metal affinity exchange resin(Bio-Rad Laboratories, Richmond, Calif.). The Chelex resinwas prepared by repeated washings in 0.5 M sodium acetateuntil a pH of 6.3 was reached and then washed five times withultrapure water (500 to 800 ml of water per wash). The Chelexresin was used at 50 or 75 g/100 ml of milk. The milk wasallowed to mix with the Chelex by stirring in a beaker overnightat 4°C and was stored refrigerated until used. The pH of thedemineralized milk was 6.0 to 6.2. The removal of metals didnot cause precipitation of proteins or other visible changes inthe milk. To add CaCl2 back to demineralized milk, the milk

TABLE 2. Metal analysis of refortified milk samples

Concn (mg/liter) in:

Metal Dialyzed milka plus:Whole milk

Chelex Chelex + Ca Chelex + Mg Chelex + Zn

P 923 516 529 470 506K 1,559 <2.5 <2.5 <2.5 <2.5Ca 1,056 9.3 667 4.2 7.8Mg 104 <0.4 <0.4 82 <0.4S 287 148 156 140 146Zn 3.4 <0.04 <0.04 <0.04 3.0Fe 0.21 0.06 0.08 <0.05 0.08Cu <0.1 <0.1 <0.1 <0.1 <0.1Na 371 739 734 665 722

a Values for dialyzed preparations were generally lower than those in Table 1because of longer dialysis and a higher concentration of Chelex (75 g/100 ml ofmilk compared with 50 g/100 ml of milk for Table 1).

VOL. 60, 1994

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 30

Dec

embe

r 20

21 b

y 14

4.13

2.16

8.21

.

APPL. ENVIRON. MICROBIOL.

was dialyzed against 20 mM sodium succinate buffer, pH 6.4, toavoid precipitation of calcium phosphate complexes.

Influence of demineralization and lysozyme addition onthermal resistance of L. monocytogenes. To determine theeffect of demineralization of whole milk on heat resistance ofL. monocytogenes, cells were added to 5 ml of whole milk ordemineralized milk at 55°C to an initial cell concentration ofapproximately 106/ml. Samples were taken every 2.5 min for upto 10 min and diluted immediately in cold PB, and survival wasdetermined by plating on TPAP. Lysozyme, when present, wasadded at a final concentration of 100 mg/liter of milk. Calcium,when present, was added at 1,100 mg/liter.

Preparation of acid-precipitated whey. The pH of wholepasteurized milk (pH 6.70) was lowered with 10 N HCl to 4.65.The milk was centrifuged at 15,000 x g for 15 min. The pH ofthe whey supernatant was raised to 6.70 with sodium hydrox-ide, and the supernatant was centrifuged again at 10,000 x gfor 10 min.

Analysis of lysozyme-treated L. monocytogenes by scanningelectron microscopy. L. monocytogenes was grown at 30°C toearly log phase in TSB (Difco), a medium in which L.monocytogenes is highly susceptible to lysozyme (25). The cellswere harvested, resuspended in PB (pH 6.2), and treated with50 or 100 mg of lysozyme per liter for 45 min at 37°C. Aftertreatment, the cells were harvested by centrifugation, fixedwith 0.25% glutaraldehyde, and captured on a 0.45-,um-pore-size membrane. The membranes were fixed on a copper gridpretreated with 0.001% polylysine for 5 min and washed withdistilled water. The cell-coated grids were washed three timeswith PB and distilled water. The cells were dehydrated withmolecular sieved ethanol in a progressive series of 30, 50, 70,80, 85, 90, 95, and 100%. The cells were super-critical-pointdried with liquid CO2 and coated with platinum to 1 nmthickness. Images were obtained with high-resolution low-voltage scanning electron microscopy at 1.5 kV. The magnifi-cation was ca. x30,000 times.

Statistical validation of the data. Each datum point in thefigures was the average of plate counts at three dilutions.Analysis of the variation of the data in the figures indicatedthat the average standard deviation for each datum point wasapproximately 14%. Because of the logarithmic plots, errorbars in most cases are smaller than the figure symbols and arenot presented. In our view, the significance of the graphs alsois not reflected in the individual datum points but rather in theprogressive trends in survival of L. monocytogenes. In severalconditions tested, the differences in viability between thecontrol and treatment groups were 4 to 6 log units, which wefeel is significant compared with the magnitude of the standarddeviation. Each experiment was replicated at least once.

RESULTS

Inactivation of L. monocytogenes by lysozyme in millk Inrepeated trials, L. monocytogenes was found to be completelyresistant to inactivation by lysozyme in whole or skim milk, butthe same batches of cells were sensitive to lysozyme in PB (Fig.1). M. luteus was rapidly lysed by lysozyme in milk (79% lysis in40 min [data not shown]), indicating that lysozyme was notinactivated in milk. Experiments were carried out to under-stand the basis of resistance of L. monocytogenes to lysozyme inmilk and to develop methods for sensitization of cells to theenzyme.Heat treatment of L. monocytogenes at 60°C for 15 s prior to

addition to milk did not significantly sensitize L. monocyto-genes to lysozyme (Fig. 1). However, preliminary experimentssuggested that sublethal heating of the cells sensitized them to

108

00

0;I-Ol

vi

D

c)0

i07

106

105

04

103

2.10

1010 50 100 150 200

Time (hours)FIG. 2. Survival of heat-treated L. monocytogenes in demineralized

milk and demineralized milk refortified with metals at 4°C. In eachcase 100 mg of lysozyme per ml and heat-treated (62.5°C for 15 s) L.monocytogenes were present. Symbols: 0, whole milk; 0, demineral-ized whole milk refortified with 667 mg of calcium per liter; O,demineralized whole milk refortified with 82 mg of magnesium perliter; A, demineralized whole milk; A, demineralized whole milkrefortified with 3.0 mg of zinc per liter.

lysozyme in buffer (data not shown; see also Fig. 3). Sincedivalent metals are known to stabilize the cell walls of gram-positive bacteria (10) and because EDTA sensitizes L. mono-cytogenes to lysozyme (16), we investigated whether the metalcontent of milk affected resistance to lysozyme. Removal ofmetals from the milk by treatment with a metal affinity resinand dialysis slightly promoted inactivation of L. monocytogenesby lysozyme (Fig. 1). When the cells were first heat treatedin PB at 60°C for 15 s and then added to demineralizedmilk, inactivation by lysozyme was dramatically increased,giving more than a 4-log-unit decrease in cell numbers over6 days at 4°C compared with control cells in milk withoutadded lysozyme. The rate of inactivation of heat-treated cellsin demineralized milk was similar to that observed in PB (Fig.1).These results suggested that minerals and mineral-associ-

ated components in milk (e.g., proteins) protect L. monocyto-genes from inactivation by lysozyme. To investigate the possiblerole of proteins, they were partially removed from milk by acidprecipitation at pH 4.65. Heat-treated L. monocytogenes wasresistant to lysozyme in acid-precipitated whey; however, thecells were sensitive in acid-precipitated whey that was subse-quently demineralized (data not shown). These results supportthe hypothesis that the metals rather than the protein compo-nents of milk protected L. monocytogenes against added ly-sozyme in milk.

3856 KIHM ET AL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 30

Dec

embe

r 20

21 b

y 14

4.13

2.16

8.21

.

INACTIVATION OF L. MONOCYTOGENES IN MILK BY LYSOZYME 3857

i 0 106

u, 10'63

0 I1I

0~~~~~~~~~~~1

~~~10'~~ ~ ~ ~~~1

102 1020 10 20 30 0 10 20 30

Time (days)FIG. 3. Survival of L. monocytogenes in 50 mM MES buffer, pH 6.4, supplemented with lysozyme (100 mg/liter) and certain metals. Unheated

L. monocytogenes (A) or L. monocytogenes heated at 62.5°C for 15 s (B) was used. Symbols: 0, control without lysozyme; *, lysozyme; A, lysozymeand calcium (1,100 mg/liter); l, lysozyme and magnesium (110 mg/liter); A, lysozyme, calcium, and magnesium.

Sensitivity of L. monocytogenes to lysozyme in metal-reforti-fied milk. Analysis of demineralized milk for metal content byinductively coupled plasma spectroscopy indicated that thelevels of Ca2", Mg2+, and Zn-+ were decreased significantlycompared with those in untreated whole milk (Table 1). Ca2+was decreased from 1,170 to 100 mg/liter, Mg2+ was decreasedfrom 118 to less than 1 mg/liter, and Zn2+ was decreased from3.5 to less than 0.1 mg/liter (Table 1). We examined whetherresistance of L. monocytogenes to lysozyme in demineralizedmilk could be restored by adding back these individual metalcomponents. The metals were refortified in demineralized milkat concentrations that would approximate their original con-centrations in whole milk. Analysis of metals in refortified milkby inductively coupled plasma spectroscopy indicated thatCa2+ was at about 63% of the original level (667 mg/liter),Mg2+ was at about 79% of the original level (82 mg/liter), andZn2+ was at about 88% of the original level (3.0 mg/liter)(Table 2). Heat-treated L. monocytogenes (62.5°C for 15 s) inrefortified milk became resistant to lysozyme when Ca2+ orMg2+, but not Zn2+, was added back to the demineralized milk(Fig. 2). Calcium supplementation restored the resistance of L.monocytogenes to a degree that approached the resistance levelin whole milk. These data support the hypothesis that themineral components of milk protect L. monocytogenes frominactivation by lysozyme.

Influence of metals on lysis of L. monocytogenes in buffer.Milk is a complex food system which contains a variety ofcomponents that may interact with the cell and protect it from

the lytic action of lysozyme. To specifically examine theinfluence of metals on the sensitivity of cells, inactivation bylysozyme in metal-free MES buffer was examined. PB was notused because of precipitation of calcium phosphate complexes.Unheated cells or cells heat treated for 15 s at 62.5°Cwere incubated in 50 mM MES containing lysozyme andCa2+ or Mg2e at 4°C. Unheated cells were only moderatelysensitive to lysozyme, and the presence of Ca2+ or Mg2+ didnot affect sensitivity (Fig. 3A). The addition of lysozymeresulted in a slow decline in cell numbers over 3 weeks at 4°C,and metal addition did not affect this slow decline. In contrast,heat-treated cells were very sensitive to lysozyme, and cellviability decreased by 4 to 6 log units over 24 days of exposure.The presence of Ca2 , Mg +, or a combination of bothprotected heat-treated L. monocytogenes from inactivationby lysozyme in buffer (Fig. 3B). For cells heated at 62.5°Cfor 15 s, the greatest viability was maintained with addedCa2' alone or the combination of added Ca2' and Mg2+ (Fig.3B). Calcium was more effective than Mg2+ in providingresistance to L. monocytogenes heated at 62.5°C for 15 s. Verysimilar results were observed with cells heated at 60°C for15 s. Control incubations of L. monocytogenes with the metaladdition but without lysozyme showed that the cells did notdie off over the course of the experiment (data not shown).These experiments support the hypothesis that inactivationwas caused by lysozyme and that Ca2+ and Mg2+ in milkprovided resistance or recovery in response to challenge bylysozyme.

VOL. 60, 1994

107,

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 30

Dec

embe

r 20

21 b

y 14

4.13

2.16

8.21

.

APPL. ENVIRON. MICROBIOL.

1071 A

106

lo05

Q

0

0

04C%Mol

rAw

1 102

10I

los,

10 2

101

1000 2 4 6 8 10 12 0 2 4 6 8 10 12

Minutes at 55°CFIG. 4. Thermal tolerance of L. monocytogenes at 55°C in whole milk or demineralized whole milk with or without lysozyme (100 mg/liter) (A)

or in demineralized milk with or without lysozyme and with or without calcium supplementation (B). (A) Symbols: 0, whole milk; 0, whole milkplus lysozyme; C1, demineralized milk without lysozyme; *, demineralized milk with lysozyme. (B) Symbols: 0, demineralized milk refortified withcalcium; 0, demineralized milk with lysozyme and refortified with calcium; A, demineralized milk; A, demineralized milk with lysozyme.

Tolerance of L. monocytogenes to thermal stress in milk. Weexamined whether the thermal sensitivity of L. monocytogeneswas altered in demineralized milk. L. monocytogenes was veryresistant to thermal inactivation at 55°C in whole milk with orwithout added lysozyme (Fig. 4A). In contrast, cell viabilitydecreased rapidly during heating in the demineralized milkcontaining lysozyme, resulting in a 6-log-unit decrease in cellnumbers after 5 min at 55°C (Fig. 4A). Interestingly, heattolerance in demineralized milk containing lysozyme appearedto be partially restored by added Ca2" (Fig. 4B). These resultsindicate that the metal composition and distribution withinmilk affects heat resistance ofL. monocytogenes in the presenceof lysozyme.

Scanning electron microscopy analysis of L. monocytogenesafter treatment with lysozyme and chelator (EDTA). Scanningelectron microscopy was performed to examine the influenceof lysozyme and EDTA (which chelates Mg2" and Ca2") onthe cell surface of L. monocytogenes Scott A. In the firstanalysis (Fig. 5), cells were grown for 16 h at 30°C in TSB.Lysozyme was added to 50 mg/liter, and the cells were incu-bated for 45 min. Lysozyme alone clearly caused digestion ofthe cell surface, producing depressions and removal of largeportions of the wall material. The partially digested cellsappeared to clump together, possibly through release of anadherent substance from the cell wall or membranes.Treatment with EDTA appeared to soften the cells and

reduce rigidity (Fig. 6b) compared with the control (Fig. 6a).Cells treated with 5 mM EDTA were slightly larger than

control cells, indicating a loosening of the cell wall rigidstructure (Fig. 6b). EDTA probably removed the metal ionsfrom the structural framework and promoted the loosening ofthe rigid cell structure. Treatment with lysozyme at 100 ppmcaused flattening and the release of globular substances fromthe cells (Fig. 6c), and this destruction was enhanced by EDTA(Fig. 6d). Cells treated with lysozyme and EDTA were exten-sively damaged as indicated by lysis and release of intracellularmaterials.

DISCUSSION

Although rigorous sanitation practices and improved qualitycontrol measures have reduced the incidence of L. monocyto-genes in many raw and minimally processed foods, the patho-gen continues to contaminate certain dairy foods, particularlyraw milk and soft-ripened cheeses (11, 30). Dairy productsappear to provide a unique environment which supportsgrowth and survival of the pathogen (26, 28). There is consid-erable interest in controlling L. monocytogenes in dairy prod-ucts, since this class of foods has been responsible for largeoutbreaks of listeriosis (11).

Postprocessing contamination is believed to be the majorroute of contamination of dairy products and some otherfoods, and it would be beneficial to incorporate antimicrobialagents that could eliminate the pathogen into foods. Lysozymeis attractive as a food preservative because it acts catalytically,thereby requiring very small quantities, and is safe in the

3858 KIHM ET AL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 30

Dec

embe

r 20

21 b

y 14

4.13

2.16

8.21

.

INACTIVATION OF L. MONOCYTOGENES IN MILK BY LYSOZYME

FIG. 5. Scanning electron microscopy analysis of L. monocytogenes treated for 45 min at 37°C as follows: (a) control; (b) lysozyme (50 mg/liter).See Materials and Methods and Results for details.

human diet. Lysozyme (mucopeptide glycohydrolase; EC3.2.1.17) catalyzes the hydrolysis of the ,B-(1-4) linkage be-tween N-acetylmuramic acid and N-acetylglucosamine resi-dues, and its specific targets are the peptidoglycan and chitincomponents in the cell walls of bacteria and fungi (21, 29). Incheeses and other fermented foods, lysozyme is primarilyactive against specific groups of gram-positive bacteria (7). Atthe concentrations generally used, egg white lysozyme does notgenerally interfere with beneficial food fermentations carriedout by lactic acid bacteria and yeasts. Egg white lysozyme hasbeen used in the European cheese industry to prevent gasformation by clostridia in hard cheeses (7) but has not yet beenapproved as generally recognized as safe (GRAS) in theUnited States. Lysozyme is usually added to the vat milk at 20to 35 mg/kg, and it associates nearly entirely with the curd,giving levels in the cheese of 200 to 400 mg/kg. The addition oflysozyme may also shorten the clotting times of milk, givinghigher yields of cheese (14). In addition to being used incheese, lysozyme has been used to prevent spoilage of variousfoods, including fresh vegetables, tofu, sausage, seafood, soysauce, and alcoholic beverages (27). Recently lysozyme hasbeen shown to have beneficial attributes in white and red wineproduction by preventing unwanted bacterial growth and as areplacement for sulfite (9).

Previous studies showed that egg white lysozyme effectivelyinactivated L. monocytogenes in vegetables but was ineffectivein Camembert cheese (17). The strong resistance of L. mono-cytogenes to inactivation by lysozyme in milk and some cheeseswas also demonstrated by Carminati and Carini (8). The basisfor this unexpected resistance in dairy products was investi-gated in this study. We have found that certain metals in milk,particularly Ca2" and Mg2", increased the resistance of L.monocytogenes to lysozyme and that sublethal heat treatmentsensitized L. monocytogenes to lytic activity.

Heat treatment ofL. monocytogenes may have disrupted wallintegrity sufficiently to allow lysozyme to reach its targetpeptidoglycan and lyse the cell. The cell surface of L. mono-cytogenes is composed of peptidoglycan, teichoic acids, andlipoteichoic acids (12, 13, 15). No outer membrane or capsuleis known to be present in the cell wall structure. The isolatedcell walls are composed of about 30 to 40% peptidoglycan,while the remainder of the dry matter is composed of teichoicand lipoteichoic acids (13) which are covalently attached to thepeptidoglycan. Whereas teichoic acids generally constitute 30to 70% of the walls of most gram-positive organisms, they aremore prevalent in Listeria spp., providing up to 75% of the drywall mass (1, 3, 4). A proposed function of the teichoic acids isthe control of cation concentration in the wall-membraneregion of the cell (2, 3). The cell walls of gram-positive bacteriacan be thought of as functioning partly as a cation exchanger,with Ca2' and Mg2+ ions having a higher affinity towards cellwalls than K+, Na+, and Li' ions but having a lesser affinitythan H+, La3+, Cd2+, or Sr2+ ions (23). Metal ions in theteichoic acid and peptidoglycan polymers have been shown tocause a contraction in cell walls, giving a more rigid structure(10, 22, 24). In the present study, treatment of L. monocyto-genes with EDTA promoted an enlargement of the cells,suggesting that removal of the metals allowed an expansion ofthe woven cell surface structure. Marquis et al. (23) reportedthat Mg2+ or Ca2+ ions pair with two anionic groups (mostlyphosphate and carboxylate groups) in peptidoglycan as well asteichoic acids in bacterial cell walls, and that laboratoryshowed that magnesium can be displaced by high concentra-tions of other cations, such as sodium. Marquis and othershave contributed to our conceptual understanding of the cellwalls of gram-positive bacteria as comprising relatively looselyknit, elastic networks of polyelectrolyte polymers. These mac-rostructures have the ability to expand and contract, depending

VOL. 60, 1994 3859

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 30

Dec

embe

r 20

21 b

y 14

4.13

2.16

8.21

.

APPL. ENVIRON. MICROBIOL.

FIG. 6. Scanning electron microscopy analysis of L. monocytogenes treated for 45 min at 37°C as follows: (a) control; (b) EDTA (5 mM); (c)lysozyme (100 mg/liter); (d) EDTA plus lysozyme. See Materials and Methods and Results for details.

on the ionic composition and pH of the medium. Sensitivity tolysozyme would therefore depend on the medium composition,particularly the presence of phosphate, organic chelators, andminerals, which would affect the porosity and charge of thegram-positive cell surface.The ionic bonds between divalent metals and anionic side

groups in teichoic acids probably contribute to protection ofthe peptidoglycan from lysozyme. Micrococcus lysodekticus,which is extremely sensitive to lysozyme, has a wall composedalmost entirely of peptidoglycan. Various chemical and physi-cal treatments may disrupt the polymeric network surroundingthe cell to a sufficient degree to allow penetration of largemolecules such as lysozyme (45 by 30 by 30 A [45 by 30 by 30x 10-10 m]). Heat treatment can disrupt the integrity, but ifenough divalent metal cations are present, the cell wall integ-rity can probably be repaired in heat-injured cells, therebyexcluding penetration by lysozyme. D-Alanine was shown to belost from cells of Staphylococcus aureus during heating, and

magnesium binding was required for repair (18). Magnesiumand iron cations were found to have a role in repair ofheat-injured L. monocytogenes (6). Our results support thehypothesis that repair of injured cells and resistance to ly-sozyme in milk are dependent on the presence of Ca2+ orMg2 . For whole bovine milk, Zhang and Allen (33) reportedthat magnesium and zinc were associated with different frac-tions: 2.3 and 1.7%, respectively, with fat, 62.0 and 8.2% withwhey, and 35.7 and 88.4% with casein. Hence, removal orpartitioning of mineral components in milk by fractionation orother technologies could influence the susceptibility of L.monocytogenes or other gram-positive pathogens to lysozymein dairy products.

It was interesting that unheated L. monocytogenes did notappear to have increased resistance to lysozyme in the pres-ence of minerals. In contrast, sublethally heated cells were ableto markedly restore lysozyme resistance and maintained via-bility in the presence of Mg2+ or Ca2+. It is possible that

3860 KIHM ET AL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 30

Dec

embe

r 20

21 b

y 14

4.13

2.16

8.21

.

INACTIVATION OF L. MONOCYTOGENES IN MILK BY LYSOZYME 3861

heating may trigger a stress response inducing the capacity torepair cell surface lesions. It has been reported that divalentcations are necessary for repair of heat-injured Listeria spp.(6). These results support the hypothesis of a dynamic inter-change of metals with the cell surface and the important rolethat metals have in promoting survival of bacteria in foods.Our data support the conclusion that heat treatment of milk

or other foods containing added lysozyme can exert a signifi-cant role in enhancing the microbiological safety of foods.Moreover, chemical and physical processing methods that arecompatible with food composition and disrupt the cell surfaceor chelate metals would be valuable in the application oflysozyme and other lytic enzymes as food preservatives.

ACKNOWLEDGMENTS

This research was supported by the International Life SciencesInstitute; by the Wisconsin Milk Marketing Board; by the Center forDairy Research, the College of Agriculture and Life Sciences, Univer-sity of Wisconsin, Madison; and by Societa Prodotti Antibiotici, Milan,Italy.We thank Ralph Albrecht and Kris Krueger for assistance with

electron microscopy.

REFERENCES1. Archibald, A. R. 1974. The structure, biosynthesis and function of

teichoic acid. Adv. Microb. Physiol. 10:53-105.2. Archibald, A. R., J. J. Armstrong, J. Baddiley, and J. B. Hay. 1961.

Teichoic acids and the structure of bacterial walls. Nature (Lon-don) 191:570-572.

3. Baddiley, J. 1972. Teichoic acids in cell walls and membranes ofbacteria. Essays Biochem. 8:35-77.

4. Baddiley, J. 1989. Bacterial cell walls and membranes. Discoveryof the techoic acids. Bioessays 10:207-210.

5. Bille, J., and M.-P. Glauser. 1988. Listeriose en Suisse. Bull.Bundesamtes Gesundheitswesen. 3:28-29.

6. Busch, S. V., and C. W. Donnelly. 1992. Development of arepair-enrichment broth for resuscitation of heat-injured Listeriamonocytogenes and Listeria innocua. Appl. Environ. Microbiol.58:14-20.

7. Carini, S., G. Mucchetti, and E. Neviani. 1985. Lysozyme: activityagainst clostridia and use in cheese production. A review. Micro-biol. Aliments Nutr. 3:299-320.

8. Carminati, D., and S. Carini. 1989. Antimicrobial activity oflysozyme against Listeria monocytogenes in milk. Microbiol. Ali-ments Nutr. 7:49-56.

9. Dell-Acqua, E. 1993. Personal communication.10. Doyle, R. J. 1989. How cell walls of gram-positive bacteria interact

with metal ions, p. 275-293. In T. J. Beveridge and R. J. Doyle(ed.), Metal ions and bacteria. John Wiley & Sons, New York.

11. Farber, J. M., and P. I. Peterkin. 1991. Listeria monocytogenes, afood-borne pathogen. Microbiol. Rev. 55:476-511.

12. Fiedler, F. 1988. Biochemistry of the cell surface of listeria strains:a locating general view. Infection 16(Suppl. 2):S92-S97.

13. Fiedler, F., and G. L. Ruhland. 1987. Structure of Listeria mono-

cytogenes cell walls. Bull. Inst. Pasteur 85:287-300.14. Giangiacomo, R., F. Nigro, G. Messina, and T. M. P. Cataneo.

1992. Lysozyme: just an additive or a technological aid as well?Food Add. Contamin. 9:427-433.

15. Hether, N. W., and L. L. Jackson. 1983. Lipotechoic acid fromListeria monocytogenes. J. Bacteriol. 156:809-817.

16. Hughey, V. L., and E. A. Johnson. 1987. Antimicrobial activity oflysozyme against bacteria involved in food spoilage and food-borne disease. Appl. Environ. Microbiol. 53:2165-2170.

17. Hughey, V. L., P. A. Wilger, and E. A. Johnson. 1989. Antibacterialactivity of hen egg white lysozyme against Listeria monocytogenesScott A in foods. Appl. Environ. Microbiol. 55:631-638.

18. Hurst, A., A. Hughes, M. Duckworth, and J. Baddiley. 1975. Lossof D-alanine during sublethal heating of Staphylococcus aureus s6and magnesium binding during repair. J. Gen. Microbiol. 89:277-284.

19. Johnson, E. A., J. H. Nelson, and M. Johnson. 1990. Microbiolog-ical safety of cheese made from heat-treated milk. II. Microbiol-ogy. J. Food Prot. 53:519-540.

20. Kihm, D. J., M. T. Hutton, J. H. Hanlin, and E. A. Johnson. 1988.Zinc stimulates sporulation in Clostridium botulinum 113B. Curr.Microbiol. 17:193-198.

21. Jolles, P., and J. Jolles. 1984. What's new in lysozyme research-always a model system, today as yesterday. Mol. Cell. Biochem.63:165-189.

22. Marquis, R. E. 1968. Salt-induced contraction of bacterial cellwalls of gram-positive cocci. J. Bacteriol. 95:775-781.

23. Marquis, R. E., K. Mayzel, and E. L. Carstensen. 1976. Cationexchange in cell walls of gram-positive bacteria. Can. J. Microbiol.22:975-982.

24. Ou, L.-T., and R. E. Marquis. 1972. Coccal wall compactness andthe swelling action of denaturants. Can. J. Microbiol. 18:623-629.

25. Premaratne, R. J., and E. A. Johnson. Unpublished data.26. Premaratne, R. J., W.-J. Lin, and E. A. Johnson. 1991. Develop-

ment of an improved chemically defined minimal medium forListeria monocytogenes. Appl. Environ. Microbiol. 57:3046-3048.

27. Proctor, V. A., and F. E. Cunningham. 1988. The chemistry oflysozyme and its use as a food preservative and a pharmaceutical.Crit. Rev. Food Sci. Nutr. 26:359-395.

28. Ryser, E. T., and E. H. Marth. 1991. Listeria, listeriosis, and foodsafety. Marcel Dekker, Inc. New York.

29. Salton, M. R J. 1957. The properties of lysozyme and its action onmicroorganisms. Bacteriol. Rev. 21:82-98.

30. Schuchat, A., K. A. Deaver, J. D. Wenger, B. D. Plikaytis, L.Mascola, R. W. Pinner, A. L. Reingold, C. V. Broome, and theListeria Study Group. 1992. Role of foods in sporadic listeriosis. I.Case-control study of dietary risk factors. J. Am. Med. Assoc.267:2041-2045.

31. Seeliger, H. P. R., and D. Jones. 1986. Genus Listeria, Pirie 1940,383AL, p. 1235-1245. In P. H. A. Sneath, N. S. Mair, M. E. Sharpe,and J. G. Holt (ed.), Bergey's manual of systematic bacteriology,vol. 2. Williams & Wilkins, Baltimore.

32. Smith, J. L., and D. L. Archer. 1988. Heat-induced injury inListeria monocytogenes. J. Ind. Microbiol. 3:105-110.

33. Zhang, P.-F., and J. C. Allen. 1993. Magnesium and zinc distribu-tions and their activity in bovine milk, p. 122. Abstr. IFT Annu.Meet., Chicago, Ill. Institute of Food Technology, Chicago.

VOL. 60, 1994

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 30

Dec

embe

r 20

21 b

y 14

4.13

2.16

8.21

.